Download Cabletron Systems EPIM F3 Specifications

Cabletron Systems
FDDI TECHNOLOGY GUIDE
notice.tmp Page i Wednesday, April 17, 1996 7:28 AM
NOTICE
Cabletron Systems reserves the right to make changes in specifications and other information
contained in this document without prior notice. The reader should in all cases consult Cabletron
Systems to determine whether any such changes have been made.
The hardware, firmware, or software described in this manual is subject to change without notice.
IN NO EVENT SHALL CABLETRON SYSTEMS BE LIABLE FOR ANY INCIDENTAL,
INDIRECT, SPECIAL, OR CONSEQUENTIAL DAMAGES WHATSOEVER (INCLUDING BUT
NOT LIMITED TO LOST PROFITS) ARISING OUT OF OR RELATED TO THIS MANUAL OR
THE INFORMATION CONTAINED IN IT, EVEN IF CABLETRON SYSTEMS HAS BEEN
ADVISED OF, KNOWN, OR SHOULD HAVE KNOWN, THE POSSIBILITY OF SUCH
DAMAGES.
 Copyright 1996 by Cabletron Systems, Inc., P.O. Box 5005, Rochester, NH 03866-5005
All Rights Reserved
Printed in the United States of America
Order Number: 9031708 April 1996
SPECTRUM, LANVIEW, Remote LANVIEW NCM-PCMMAC, MicroMMAC, and BRIM are
registered trademarks and Element Manager, EPIM, EPIM-A, EPIM-F1, EPIM-F2, EPIM-F3,
EPIM-T, EPIM-X, FOT-F, FOT-F3, HubSTACK, SEH, SEHI, and TMS-3 are trademarks of
Cabletron Systems, Inc.
All other product names mentioned in this manual may be trademarks or registered trademarks of
their respective companies.
Printed on
FDDI Technology Manual
Recycled Paper
i
notice.tmp Page ii Wednesday, April 17, 1996 7:28 AM
Notice
FCC NOTICE
This device complies with Part 15 of the FCC rules. Operation is subject to the following two
conditions: (1) this device may not cause harmful interference, and (2) this device must accept any
interference received, including interference that may cause undesired operation.
NOTE: This equipment has been tested and found to comply with the limits for a Class A digital
device, pursuant to Part 15 of the FCC rules. These limits are designed to provide reasonable
protection against harmful interference when the equipment is operated in a commercial environment.
This equipment uses, generates, and can radiate radio frequency energy and if not installed in
accordance with the operator’s manual, may cause harmful interference to radio communications.
Operation of this equipment in a residential area is likely to cause interference in which case the user
will be required to correct the interference at his own expense.
WARNING: Changes or modifications made to this device which are not expressly approved by the
party responsible for compliance could void the user’s authority to operate the equipment.
DOC NOTICE
This digital apparatus does not exceed the Class A limits for radio noise emissions from digital
apparatus set out in the Radio Interference Regulations of the Canadian Department of
Communications.
Le présent appareil numérique n’émet pas de bruits radioélectriques dépassant les limites applicables
aux appareils numériques de la class A prescrites dans le Règlement sur le brouillage radioélectrique
édicté par le ministère des Communications du Canada.
VCCI NOTICE
This equipment is in the 1st Class Category (information equipment to be used in commercial and/or
industrial areas) and conforms to the standards set by the Voluntary Control Council for Interference
by Information Technology Equipment (VCCI) aimed at preventing radio interference in commercial
and/or industrial areas.
Consequently, when used in a residential area or in an adjacent area thereto, radio interference may be
caused to radios and TV receivers, etc.
Read the instructions for correct handling.
ii
FDDI Technology Manual
notice.tmp Page iii Wednesday, April 17, 1996 7:28 AM
Notice
CABLETRON SYSTEMS, INC. PROGRAM LICENSE AGREEMENT
IMPORTANT: Before utilizing this product, carefully read this License Agreement.
This document is an agreement between you, the end user, and Cabletron Systems, Inc. (“Cabletron”)
that sets forth your rights and obligations with respect to the Cabletron software program (the
“Program”) contained in this package. The Program may be contained in firmware, chips or other
media. BY UTILIZING THE ENCLOSED PRODUCT, YOU ARE AGREEING TO BECOME
BOUND BY THE TERMS OF THIS AGREEMENT, WHICH INCLUDES THE LICENSE AND
THE LIMITATION OF WARRANTY AND DISCLAIMER OF LIABILITY. IF YOU DO NOT
AGREE TO THE TERMS OF THIS AGREEMENT, PROMPTLY RETURN THE UNUSED
PRODUCT TO THE PLACE OF PURCHASE FOR A FULL REFUND.
CABLETRON SOFTWARE PROGRAM LICENSE
1.
LICENSE. You have the right to use only the one (1) copy of the Program provided in this
package subject to the terms and conditions of this License Agreement.
You may not copy, reproduce or transmit any part of the Program except as permitted by the
Copyright Act of the United States or as authorized in writing by Cabletron.
2.
OTHER RESTRICTIONS. You may not reverse engineer, decompile, or disassemble the
Program.
3.
APPLICABLE LAW. This License Agreement shall be interpreted and governed under the laws
and in the state and federal courts of New Hampshire. You accept the personal jurisdiction and
venue of the New Hampshire courts.
EXCLUSION OF WARRANTY AND DISCLAIMER OF LIABILITY
1.
EXCLUSION OF WARRANTY. Except as may be specifically provided by Cabletron in
writing, Cabletron makes no warranty, expressed or implied, concerning the Program (including
its documentation and media).
CABLETRON DISCLAIMS ALL WARRANTIES, OTHER THAN THOSE SUPPLIED TO
YOU BY CABLETRON IN WRITING, EITHER EXPRESSED OR IMPLIED, INCLUDING
BUT NOT LIMITED TO IMPLIED WARRANTIES OF MERCHANTABILITY AND
FITNESS FOR A PARTICULAR PURPOSE, WITH RESPECT TO THE PROGRAM, THE
ACCOMPANYING WRITTEN MATERIALS, AND ANY ACCOMPANYING HARDWARE.
2.
NO LIABILITY FOR CONSEQUENTIAL DAMAGES. IN NO EVENT SHALL
CABLETRON OR ITS SUPPLIERS BE LIABLE FOR ANY DAMAGES WHATSOEVER
(INCLUDING, WITHOUT LIMITATION, DAMAGES FOR LOSS OF BUSINESS,
PROFITS, BUSINESS INTERRUPTION, LOSS OF BUSINESS INFORMATION, SPECIAL,
INCIDENTAL, CONSEQUENTIAL, OR RELIANCE DAMAGES, OR OTHER LOSS)
ARISING OUT OF THE USE OR INABILITY TO USE THIS CABLETRON PRODUCT,
EVEN IF CABLETRON HAS BEEN ADVISED OF THE POSSIBILITY OF SUCH
DAMAGES. BECAUSE SOME STATES DO NOT ALLOW THE EXCLUSION OR
LIMITATION OF LIABILITY FOR CONSEQUENTIAL OR INCIDENTAL DAMAGES, OR
ON THE DURATION OR LIMITATION OF IMPLIED WARRANTIES, IN SOME
INSTANCES THE ABOVE LIMITATIONS AND EXCLUSIONS MAY NOT APPLY TO
YOU.
FDDI Technology Manual
iii
notice.tmp Page iv Wednesday, April 17, 1996 7:28 AM
Notice
UNITED STATES GOVERNMENT RESTRICTED RIGHTS
The enclosed product (a) was developed solely at private expense; (b) contains “restricted computer
software” submitted with restricted rights in accordance with Section 52227-19 (a) through (d) of the
Commercial Computer Software - Restricted Rights Clause and its successors, and (c) in all respects
is proprietary data belonging to Cabletron and/or its suppliers.
For Department of Defense units, the product is licensed with “Restricted Rights” as defined in the
DoD Supplement to the Federal Acquisition Regulations, Section 52.227-7013 (c) (1) (ii) and its
successors, and use, duplication, disclosure by the Government is subject to restrictions as set forth in
subparagraph (c) (1) (ii) of the Rights in Technical Data and Computer Software clause at
252.227-7013. Cabletron Systems, Inc., 35 Industrial Way, Rochester, New Hampshire 03867-0505.
iv
FDDI Technology Manual
Contents
Overview
PURPOSE OF THIS MANUAL...................................................................................... v
WHO SHOULD USE THIS MANUAL ......................................................................... v
STRUCTURE OF THIS MANUAL ................................................................................ v
RELATED DOCUMENTS.............................................................................................. vi
Chapter 1
INTRODUCTION
FDDI OVERVIEW ......................................................................................................... 1-1
FDDI FEATURES .......................................................................................................... 1-2
Bandwidth............................................................................................................... 1-2
Transmission Medium........................................................................................... 1-2
Fault Recovery........................................................................................................ 1-3
Frame Transmission............................................................................................... 1-3
Media Access Method ........................................................................................... 1-3
Chapter 2
FDDI DEVICES
FDDI STATIONS ........................................................................................................... 2-2
FDDI CONCENTRATORS........................................................................................... 2-3
FDDI BRIDGES.............................................................................................................. 2-4
OPTICAL BYPASS SWITCHES................................................................................... 2-5
Chapter 3
FDDI PHYSICAL CONNECTIONS
FDDI FIBER CONNECTORS ...................................................................................... 3-1
MIC Connector Ports............................................................................................. 3-2
FDDI TWISTED PAIR CONNECTORS ..................................................................... 3-3
Twisted Pair Port Pinouts ..................................................................................... 3-3
FDDI PORT CONNECTION RULES ......................................................................... 3-4
Chapter 4
FDDI FRAME FORMATS
FDDI DATA FRAMES .................................................................................................. 4-1
TOKEN FRAMES .......................................................................................................... 4-3
i
Chapter 5
FDDI RING TOPOLOGY
DUAL RING WITHOUT TREES................................................................................. 5-1
DUAL RING WITH TREES ......................................................................................... 5-2
WRAPPED RING .......................................................................................................... 5-3
SINGLE TREE ................................................................................................................ 5-4
DUAL HOMING ........................................................................................................... 5-4
Chapter 6
FDDI RING OPERATION
STATION INITIALIZATION....................................................................................... 6-1
RING INITIALIZATION.............................................................................................. 6-4
Transmitting the First Token................................................................................. 6-6
Ring Initialization................................................................................................... 6-6
Synchronous Transmission ................................................................................... 6-7
Asynchronous Transmission ................................................................................ 6-7
Token Priorities....................................................................................................... 6-7
Restricted/Non-Restricted Token Mode ............................................................ 6-8
Ring Timing and Latency...................................................................................... 6-8
Maximum Ring Latency........................................................................................ 6-8
Total Transmission Time ....................................................................................... 6-9
Token Rotation Timer ............................................................................................ 6-9
Token Hold Timer .................................................................................................. 6-9
BASIC RING OPERATION.......................................................................................... 6-9
The Beacon Process .............................................................................................. 6-10
Frame Transmission............................................................................................. 6-10
Ring Fault Recovery ............................................................................................ 6-11
OPTICAL BYPASS SWITCH .............................................................................. 6-12
Chapter 7
BRIDGING WITH THE FDMMIM
ETHERNET FRAME TYPES........................................................................................ 7-1
Ethernet II Frame Type.......................................................................................... 7-3
Ethernet “Raw” Frame Type ................................................................................ 7-3
Ethernet 802.2 Frame Type.................................................................................... 7-4
Ethernet SNAP Frame Type.................................................................................. 7-5
FDDI FRAME TYPES.................................................................................................... 7-6
FDDI 802.2 Frame Type......................................................................................... 7-7
FDDI SNAP Frame Type ....................................................................................... 7-8
ETHERNET TO FDDI BRIDGING.............................................................................. 7-9
Ethernet II to FDDI SNAP Frame Bridging...................................................... 7-10
802.3 “Raw” Frame to FDDI MAC Frame Bridging ....................................... 7-10
802.2 Frame To FDDI 802.2 Frame Bridging .................................................... 7-10
Ethernet SNAP Frame to FDDI SNAP Frame Bridging ................................. 7-11
FDDI TO ETHERNET BRIDGING............................................................................ 7-11
Case 1.............................................................................................................. 7-12
Case 2.............................................................................................................. 7-12
Case 3.............................................................................................................. 7-12
Case 4.............................................................................................................. 7-13
ii
Appendix A ANSI STANDARDS FOR FDDI
THE OSI NETWORK MODEL................................................................................... A-2
STATION MANAGEMENT (SMT) ........................................................................... A-3
SMT Frame Services ............................................................................................. A-3
Connection Management..................................................................................... A-4
Ring Management (RMT) .................................................................................... A-5
MEDIA ACCESS CONTROL (MAC) ........................................................................ A-6
PHYSICAL LAYER PROTOCOL (PHY) ................................................................... A-6
4-Bit/5-Bit Encoding/Decoding Scheme .......................................................... A-7
Clock Synchronization ......................................................................................... A-7
Elasticity Buffer ..................................................................................................... A-7
PHYSICAL MEDIUM DEPENDANT (PMD) .......................................................... A-8
Appendix B FDDI SPECIFICATIONS
FDDI DESIGN CONSIDERATIONS ..........................................................................B-2
Ring Length ............................................................................................................B-2
Drive Distance ........................................................................................................B-2
Attenuation .............................................................................................................B-3
Number of Stations................................................................................................B-3
iii
PREFACE
PURPOSE OF THIS MANUAL
Welcome to the FDDI Technology Manual. This manual provides a basic overview
of Fiber Distributed Data Interface (FDDI) technology. The objective of this
manual is to help Cabletron’s customers better understand FDDI concepts and
network operation.
WHO SHOULD USE THIS MANUAL
This manual is intended for users of Cabletron’s FDDI products and should be
used as a supplement to Cabletron’s FDDI User Manuals.
STRUCTURE OF THIS MANUAL
This manual is organized as follows:
Chapter 1, Introduction - Introduces FDDI features and characteristics.
Chapter 2, FDDI Devices - Describes FDDI devices and their functions.
Chapter 3, FDDI Physical Connections - Describes how FDDI devices physically
attach to the ring.
Chapter 4, FDDI Frame Formats - Describes FDDI Frame and Token formats.
Chapter 5, FDDI Ring Topology - Outlines various FDDI ring topologies.
Chapter 6, FDDI Ring Operation - Outlines basic FDDI ring operation including
station initialization, the claim token process and basic ring operation.
Chapter 7, Bridging with the FDMMIM- Provides a basic overview of bridging
with Cabletron’s FDMMIM.
Appendix A, ANSI Standards for FDDI - Describes the FDDI standards outlined
by the ANSI standards committee.
Appendix B, FDDI Rules and Specifications - Provides a quick reference for FDDI
specifications and network requirements.
v
PREFACE
RELATED DOCUMENTS
The American National Standards Institute (ANSI) Accredited Standards
Committee (ASC) Task Group X3T9.5 writes all FDDI standards. The ANSI
committee consists of representatives from various networking companies.
Cabletron is an active member of the ANSI committee and strictly adheres to
these standards while designing new products. For additional FDDI information,
refer to the following ANSI documents:
•
Station Management (SMT) - ANSI X3.229
•
Media Access Control (MAC) - ANSI X3.139-1987
•
Physical Layer Protocol (PHY) - ANSI X3.148-1988
•
Multimode Fiber Physical Layer Medium Dependant (PMD) - ANSI X3.166
•
Single Mode Fiber Physical Layer Medium Dependent (SMF-PMD) - ANSI
X3.184
•
Twisted Pair Physical Layer Medium Dependent (TP-PMD) - ANSI
X3T9.5/94-044
Each document describes essential FDDI entities. These entities perform tasks
which are essential to the operation of the FDDI network including media access,
token passing, and frame generation.
vi
Chapter 1
INTRODUCTION
This chapter introduces Fiber Distributed Data Interface (FDDI) features and
describes characteristics that distinguish FDDI from other Local Area Network
(LAN) technologies such as Ethernet and Token Ring.
FDDI OVERVIEW
FDDI is a 100 megabit per second LAN technology that transmits data frames
over dual counter-rotating rings. It is an ideal network backbone technology and
is typically used to connect lower speed LANs such as Ethernet—10 megabits per
second and Token Ring—4/16 megabits per second. FDDI is primarily a fiber
optic network that was originally designed to operate over multimode fiber optic
cable but has been modified to operate over single mode fiber optic cable,
unshielded twisted pair cable, and shielded twisted pair cable. Figure 1-1 is an
example of an FDDI network and shows some of its distinguishing features and
components.
SINGLE ATTACHED
STATIONS
PRIMARY RING
DUAL ATTACHED
CONCENTRATOR
SECONDARY RING
FDDI DUAL
COUNTER-ROTATING RING
NETWORK
DUAL ATTACHED
CONCENTRATOR
SINGLE ATTACHED
CONCENTRATOR
Figure 1-1. Typical FDDI Network
1-1
INTRODUCTION
FDDI FEATURES
FDDI’s most distinguishing features are directly related to the fiber optic
medium. Although twisted pair cable is a valid FDDI transmission medium, it
does not match the performance features of fiber and is used primarily as a low
cost solution for desktop connections. The fiber optic medium provides a number
of advantages over twisted pair including greater transmission distances, fault
recovery, and security. The following definitions highlight some of FDDI’s most
important features.
Bandwidth
FDDI bandwidth is 100 megabits per second which is considerably faster than
Ethernet—10 megabits per second or Token Ring—4 or/16 megabits per second.
The increased bandwidth of FDDI is ideal for the transmission of voice, video, or
data.
Transmission Medium
FDDI transmits data frames over a physical medium of multimode fiber optic
cable, single mode fiber optic cable, unshielded twisted pair cable, and shielded
twisted pair cable. Multimode fiber optic cable and single mode fiber optic cable
provide backbone ring connections while unshielded and shielded twisted pair
cable provide low cost connections from the fiber backbone to the desktop.
Fiber optic cable provides a number of advantages over copper wire, including:
1-2
•
Ring Length: The maximum ring length for a dual-ring FDDI network is 100
Kilometers (60 Miles). The maximum ring length for an FDDI network in the
wrapped state (single ring) is 200 Kilometers (120 Miles).
•
Number of Stations per Network: FDDI allows up to 500 stations per
network.
•
Security: Fiber cables cannot be tapped without disruption to the ring.
•
Immunity from Electromagnetic Interference: Fiber is not affected by
electromagnetic interference.
FDDI FEATURES
Fault Recovery
FDDI has a dual counter-rotating ring topology that provides a primary path for
normal operation and a secondary path for fault recovery. If the primary ring
fails, FDDI changes the data path to the secondary ring.
Frame Transmission
FDDI stations communicate on the ring using the following message formats:
•
Frames: provide information concerning ring management, network
problems, and statistics.
•
Token: The token is a special frame that controls access to the ring. Only the
FDDI station holding the token can transmit data on the ring.
Refer to Chapter 4, FDDI Frames for more information about frames.
Media Access Method
FDDI uses a token passing media access method to transmit frames on the ring. A
token is a special frame that circulates around the ring. When an FDDI station
needs to transmit data, it captures the token, transmits the data frames, and
reissues the token. Only the FDDI station that possesses the token can transmit
data. Chapter 6, FDDI Ring Operation describes the claim token procedure.
1-3
Chapter 2
FDDI DEVICES
This chapter describes devices that are common to an FDDI network. All devices
attached to an FDDI ring must comply with The ANSI X3T9.5 Standards outlined
in Appendix A. Typical FDDI devices include stations, concentrators, bridges.
and optical bypass switches. Figure 2-1 shows various devices attached to an
FDDI ring. The following sections provide a description of each of these devices
and their network functions.
SINGLE ATTACHED
STATIONS
ETHERNET NETWORK
FDDI TO
ETHERNET
BRIDGE
PRIMARY RING
DUAL ATTACHED
CONCENTRATOR
FDDI DUAL
COUNTER-ROTATING RING
NETWORK
SECONDARY RING
DUAL ATTACHED
STATION
(FILE SERVER)
DUAL ATTACHED
CONCENTRATOR
SINGLE ATTACHED
CONCENTRATOR
Figure 2-1. Devices on an FDDI Ring
2-1
FDDI DEVICES
FDDI STATIONS
FDDI stations are addressable nodes on an FDDI network capable of transmitting,
receiving, and repeating information. Workstations, Fileservers, and Printers are
examples of FDDI stations. Stations connect to the ring using one of the following
configurations:
•
Single-Attachment Station (SAS): connects to one ring.
•
Dual-Attachment Station (DAS): connects to both rings.
Figure 2-2 shows each of the station configurations.
SINGLE ATTACHED
STATION
DUAL ATTACHED
STATION
MAC
MAC
MAC
PHY
PMD
SMT
SMT
In
Out
Primary
In
PHY-A
PHY-B
PMD-A
PMD-B
Secondary
Out
Secondary
In
Figure 2-2. FDDI Stations
2-2
Primary
Out
FDDI CONCENTRATORS
FDDI CONCENTRATORS
FDDI concentrators are central hubs that provide connections to the ring for
single attached stations. Concentrators may or may not have a MAC entity and
connect to the ring using one of the following configurations:
•
Null-Attachment Concentrator - Does not connect to the Dual Rings.
•
Single-Attachment Concentrator (SAC) - connects to one ring.
•
Dual-Attachment Concentrator (DAC) - connects to both rings.
Figure 2-3 shows each of the FDDI concentrator configurations:
DUAL ATTACHED
CONCENTRATOR
SINGLE ATTACHED
CONCENTRATOR
PMD-3
PMD-2
PMD-1
PHY-3
PHY-2
PHY-1
MAC
MAC
PMD-3
PMD-2
PMD-1
PHY-3
PHY-2
PHY-1
MAC
PHY
SMT
PMD
SMT
In
Out
Primary
In
PHY-A
PHY-B
PMD-A
PMD-B
Secondary
Out
Secondary
In
Primary
Out
Figure 2-3. FDDI Concentrators
2-3
FDDI DEVICES
FDDI BRIDGES
FDDI bridges connect multiple FDDI networks. They also link FDDI rings to
similar networks such as Token Ring or Ethernet. Similar networks have the same
upper five layers of the OSI model but have different Link and Physical layers. A
bridge does not expand an existing FDDI ring, it connects rings. Figure 2-4 is an
example of an FDDI bridge configuration.
ETHERNET NETWORK
FDDI TO
ETHERNET
BRIDGE
FDDI DUAL
COUNTER-ROTATING RING
NETWORK
Figure 2-4. FDDI Bridge
Bridges typically use one of two bridging techniques, encapsulation or
translation. Translation bridges translate non-FDDI MAC layer packets to FDDI
data frames. For example, translation bridges allow an Ethernet station to talk to
an FDDI station.
Encapsulation bridges enclose the non-FDDI packets within the FDDI protocol
and therefore must be installed in pairs. The sending bridge encapsulates the
message and the receiving bridge strips the FDDI protocol, restoring the original
frame. The bridge maintains routing information used to filter (prevent frames
from crossing the bridge) or forward messages across the bridge.
2-4
OPTICAL BYPASS SWITCHES
OPTICAL BYPASS SWITCHES
Optical bypass switches maintain ring continuity if an FDDI station fails or
becomes removed from the ring. This device is inserted between a station and the
FDDI ring connection and provides passive optical switching of both the primary
and secondary ring cables. For example, if an FDDI station fails, the optical
bypass switch automatically diverts FDDI frames through the switch instead of
through the station. This prevents a wrap condition on the FDDI ring.
Figure 2-5 shows an optical bypass switch and the data paths through the switch
in both the bypass and operational (non-bypassed) states.
NOTE
Optical bypass switches do not connect Singlemode fiber connections to Multimode fiber
connections.
BYPASS STATE
OPERATIONAL STATE
Station Power Off
Station Power On
Station
Station
FDDI Dual
Optical Bypass
Switch
FDDI Dual
Optical Bypass
Switch
FDDI
Ring
FDDI
Ring
Figure 2-5. Optical Bypass Switch
2-5
Chapter 3
FDDI PHYSICAL CONNECTIONS
This chapter describes FDDI connector types and FDDI connection rules.
Multimode fiber and single mode fiber optic cable use Media Interface
Connectors (MICs) to attach to FDDI ports while Twisted pair cable attaches to
concentrators using RJ45 ports and connectors. The following section describe
each connector type.
FDDI FIBER CONNECTORS
Multimode fiber and singlemode fiber optic cable use Media Interface Connectors
(MICs) to attach to FDDI ports. The MIC consists of two halves: a connector and a
receptacle. The connector is the male half, which resides on the fiber optic cable.
The receptacle is the female half, which resides on the FDDI station. Both the
connector and receptacle have keys which ensure proper alignment of the
primary and secondary fibers. Figure 3-1 shows each MIC configuration.
PRIMARY IN
SECONDARY OUT
Type A
TYPE A
DUAL
ATTACHMENT
SECONDARY IN
PRIMARY OUT
Type B
TYPE B
IN
OUT
Type M
TYPE M
SINGLE
ATTACHMENT
IN
OUT
Type S
TYPE S
Figure 3-1. Fiber Optic Connectors and Receptacles
3-1
FDDI PHYSICAL CONNECTIONS
MIC Connector Ports
MIC connectors have keys which distinguish port types. Types A, B, and M are
precision connectors, mechanically keyed to ensure proper connections to
Primary Ring-In and Primary Ring-Out fibers. The Type S connector has a wide,
centrally located key and is considered a non-precision connector for use at the
station end of a Single Attachment Station lobe cable. The following list describes
each port’s function on the FDDI ring:
3-2
•
A ports - receive data from the Primary ring, and transmit data to the
Secondary ring. Type A ports provide dual attachment to the primary and
secondary data paths of the main ring.
•
B ports - receive data from the Secondary ring, and transmit data to the
Primary ring. Type B ports provide dual attachment to the primary and
secondary data paths of the main ring.
•
M (Master) ports - receive and transmit data from same ring. Type M ports are
used for single attachment stations and concentrators.
•
S (Slave) ports - receive and transmit data from same ring. Type S ports are
used for single attachment stations and concentrators.
FDDI TWISTED PAIR CONNECTORS
FDDI TWISTED PAIR CONNECTORS
The Twisted Pair Physical Layer Medium Dependent (TP-PMD) ANSI
specification is a working draft and many of the standards proposed by the
TP-PMD have not been approved by the networking industry. The twisted pair
specifications listed in this document are specific to Cabletron only.
Cabletron’s FDDI products use twisted pair cabling to connect Single Attached
Stations to FDDI concentrators. Twisted pair cable does not replace fiber cabling
on the dual backbone ring. Twisted pair configurations use the following port
types:
•
M (Master) ports - receive and transmit data from same ring. Type M ports are
used for single attachment stations and concentrators.
•
S (Slave) ports - receive and transmit data from same ring. Type S ports are
used for single attachment stations and concentrators.
Twisted Pair Port Pinouts
To connect a Type M concentrator port to a Type S station port the twisted pair
cable must be crossed over. Figure 3-3 shows the pinouts for a twisted pair port.
Pin 1
RJ-45 TP-PMD PORT
Contact
Signal
1
Transmit +
2
Transmit —
3
N/A
4
N/A
5
N/A
6
N/A
7
Receive +
8
Receive —
Caution: Ground only one end of an
STP segment. For Cabletron TP-PMD
products, the port casing is grounded.
Figure 3-2. Pinouts for an FDDI RJ45 Port
3-3
FDDI PHYSICAL CONNECTIONS
FDDI PORT CONNECTION RULES
In a typical FDDI ring, the following rules apply:
•
A ports should only connect to B ports.
•
B ports should only connect to A ports.
•
M ports should only connect to S ports.
All other port-to-port connections are either Illegal or Undesirable because they
may result in unexpected ring topologies. The Station Management entity checks
for Illegal or Undesirable connections when any link is established. If the
connection is Illegal, then the connection is automatically dropped. If the
connection is Undesirable, allowance of the connection is up to the Connection
Policy of connection nodes.
A primary functions of the Station Management entity is to control physical
connections among A, B, M, and S type ports. Table 3-1 summarizes the FDDI
connection rules.
Table 3-1. FDDI Connection Rules
Port 2
3-4
A
B
S
M
P
o
r
t
A
V,U
V
V,U
V,P
B
V
V,U
V,U
V,P
S
V,U
V,U
V
V
1
M
V
V
V
X
•
V - valid connection
•
X - illegal connection
•
U - undesirable connection
•
P - valid, but when both A and B are connected to M ports, only the B
connection is used. Connecting A and B to M ports creates a dual homing
configuration. Dual homing is a method of configuring concentrators with a
redundant topology.
Chapter 4
FDDI FRAME FORMATS
This chapter describes FDDI frame formats. The MAC entity generates two basic
message formats, Tokens and Frames. The following sections describe each
message format.
FDDI DATA FRAMES
Figure 4-1 shows the overall format of an FDDI Token and Data frame:
TOKEN
Preamble
≥ 16 Symbols
Starting
Delimiter
2 Symbols
J K
FRAME
Preamble
≥ 16 Symbols
Starting
Delimiter
2 Symbols
Frame
Control
2 Symbols
Ending
Delimiter
2 Symbols
T T
Frame Check
Sequence Coverage
Frame
Control Destination Address
2 Symbols 4 or 12 Symbols
Source Address
4 or 12 Symbols
Information
≥ 0 Symbol Pairs
T
Frame Check Ending
Delimiter
Sequence
8 Symbols 1 Symbol
Frame Status
≥ 3 Symbols
Maximum - 9000 symbols
Figure 4-1. FDDI Frame Formats
4-1
FDDI FRAME FORMATS
Table 4-1. FDDI Data Frame Layout
Field Name
Field Definition
Preamble
16 + symbols
Signals the start of a valid frame.
Start Delimiter
2 Symbols
Signals that FC is next field.
Frame Control
2 Symbols
Identifies the type of frame (MAC, LLC, etc.).
Destination Address
4 or 12
Symbols
Address of the destination of the packet.
Source Address
4 or 12
Symbols
Address of the packets origin.
Information (Data)
≤ 8956
Symbols
Contains the data to be transferred.
FCS (Frame Check
Sequence)
8 Symbols
Used to determine integrity of the packet.
ED (Ending
Delimiter)
1 Symbol
Signals the end of the frame.
FS (Frame Status)
3 Symbols
Indicates the status of the frame.
NOTE
4-2
Field Size
FDDI uses a 5 bit symbol scheme. The PHY handles the encoding and decoding of the
four bit to five bit symbols.
TOKEN FRAMES
TOKEN FRAMES
The Token is made up of 22 symbols. Table 4-2 explains each of the Token Frame
fields.
Table 4-2. FDDI Frame Type
Field Name
Field Size
Field Definition
Preamble
16 + symbols
Maintains clock synchronization
Start
Delimiter
2 Symbols
Signals the start of a valid token.
Frame
Control
2 Symbols
Distinguishes a Token from a Frame.
ED (Ending
Delimiter)
1 Symbol
Signals the end of the Token.
4-3
Chapter 5
FDDI RING TOPOLOGY
This chapter describes FDDI ring topologies as well as FDDI design
considerations that may be useful to a network designer.
DUAL RING WITHOUT TREES
The Dual Ring Without Trees configuration consists of dual attachment stations
that form a ring by connecting A ports to B ports and B ports A ports. This
configuration is commonly used in small engineering/research environments to
localize FDDI rings. The disadvantage of the Dual Ring Without Trees
configuration is the risk of a wrap condition in the event of a station failure.
Optical bypass switches are typically used in this configuration to prevent a wrap
condition. Figure 5-1 shows a Dual Ring Without Trees configuration.
PRIMARY RING
DUAL ATTACHED
STATION
SECONDARY RING
FDDI DUAL
COUNTER-ROTATING RING
NETWORK
DUAL ATTACHED
STATION
Figure 5-1. FDDI Dual Ring Topology
5-1
FDDI RING TOPOLOGY
DUAL RING WITH TREES
A Dual Ring with Trees configuration uses dual attachment concentrators, single
attachment concentrators, and single attachment stations to form tree structures.
Single attachment stations and single attachment concentrators connect to the
dual attachment concentrator M ports instead of the ring. This configuration
enhances network reliability because dual attachment concentrators
automatically reconfigure the network each time stations are inserted or removed
from the ring.
Another advantage of the Dual Ring with Trees configuration is cost. Adapter
boards for single attachment stations are cheaper than adapter boards for dual
attachment concentrators. Figure 5-2 shows a typical Dual Ring with Trees
configuration.
SINGLE ATTACHED
STATIONS
PRIMARY RING
DUAL ATTACHED
CONCENTRATOR
SECONDARY RING
FDDI DUAL
COUNTER-ROTATING RING
NETWORK
DUAL ATTACHED
CONCENTRATOR
SINGLE ATTACHED
STATIONS
SINGLE ATTACHED
CONCENTRATOR
Figure 5-2. FDDI Dual Ring of Trees
5-2
TREE STRUCTURE
WRAPPED RING
WRAPPED RING
A Wrapped Ring is the result of a broken cable or a faulty Dual Attachment
Concentrator or Dual Attachment Station. Figure 5-3 shows a cut cable between
concentrator 3 (downstream neighbor) and concentrator 4 (upstream neighbor).
Both Concentrator 3 and Concentrator 4 wrap. This scenario repeats for a failed
station or concentrator. Both the upstream and downstream neighbors wrap.
When a Dual Attachment Concentrator or Dual Attachment Station wraps a port,
it internally connects the primary ring to the secondary ring. This maintains a
data path for frame transmission by creating one contiguous enveloped ring.
Similarly, if a Dual Attachment Station or Dual Attachment Concentrator is
powered off or removed from the ring, the upstream and downstream neighbors
wrap.
RING WRAP
DUAL ATTACHED
CONCENTRATOR 3
CABLE FAILURE
PRIMARY RING
A
B
DUAL ATTACHED
CONCENTRATOR 4
SECONDARY RING
A
B
A
DUAL ATTACHED
CONCENTRATOR 2
FDDI DUAL
COUNTER-ROTATING RING
NETWORK
B
B
A
DUAL ATTACHED
CONCENTRATOR 1
Figure 5-3. Wrapping a Broken Ring
5-3
FDDI RING TOPOLOGY
SINGLE TREE
The Single Tree topology does not use the dual ring, only the single ring. All of the
network devices are single attachment stations and single attachment
concentrators. Since the Single Tree topology only uses the single ring, there is no
back up path if a cable fails. If an individual single attachment station fails, it does
not affect the rest of the network. If a single attachment concentrator fails, the
single attachment stations become separated from the rest of the network.
Figure 5-4 shows a Single Tree topology.
SINGLE ATTACHED
STATIONS
SINGLE ATTACHED
CONCENTRATOR
Figure 5-4. FDDI Tree Topology
DUAL HOMING
Dual Homing provides redundant paths to critical network devices such as
fileservers, bridges or workstations. This configuration requires the following:
•
The FDDI network must be a Dual Ring With Trees.
•
The critical device must be a Dual Attachment Station.
•
The critical device must attach to Dual Attachment Concentrators.
Figure 5-5 illustrates a Dual Homing configuration. The Dual Attachment Station
connects to the M ports of two different Dual Attachment Concentrators. FDDI
connection rules considers this an Undesirable connection, but the Station
Management entity allows the B to M port connection to be active while the A to
M port connection remains in standby. This creates a redundant data path for the
critical network device. If for any reason the B to M connection is lost, the Station
Management entity switches the data path to the A to M connection. This protects
the critical network device from a failure on either Concentrator.
5-4
DUAL HOMING
FDDI DUAL ATTACHED
STATION
FILESERVER
DUAL ATTACHED
CONCENTRATOR 3
DUAL HOMED
CONNECTION
PRIMARY RING
SECONDARY RING
FDDI DUAL
COUNTER-ROTATING RING
NETWORK
DUAL ATTACHED
CONCENTRATOR 4
DUAL ATTACHED
CONCENTRATOR 2
DUAL ATTACHED
CONCENTRATOR 1
Figure 5-5. Dual Homing Topology
5-5
Chapter 6
FDDI RING OPERATION
This chapter describes basic FDDI ring operation, including:
•
Station Initialization
•
Ring Initialization
Proper operation of the FDDI ring requires interaction between the Station
Management (STM), Media Access Control (MAC), Physical Layer Protocol
(PHY), and Physical Layer Media Dependent (PMD) entities. The Station
Management entity is responsible for coordinating station initialization and the
normal operation of the FDDI ring. Refer to Appendix A for more information
concerning FDDI entities.
STATION INITIALIZATION
When a station attaches to the ring, it must run an initialization procedure. The
station initialization procedure checks the integrity of the fiber optic link and
determines if the ports on each station are ready to exchange data. Station
initialization is a function of Physical Connection Management (PCM), which is a
subentity of Station Management.
Figure 6-1 shows a new station attaching to the ring. The Station Initialization
procedure begins when the new station sends signals to the PCM entity of the
downstream station. The PCM entities of each station then begins the station
initialization procedure, which begins when the Physical Layer Protocol detects
an active fiber optic link. The following sections explain each test.
6-1
FDDI RING OPERATION
STATION C
STATION B
Link B/C
PHY B
PHY B
MAC
Link C/A
MAC
Link A/B
PHY A
PHY A
PHY A
PHY B
MAC
STATION A
Figure 6-1. FDDI Station Initialization
6-2
STATION INITIALIZATION
Table 6-1 explains the station initialization procedure.
Table 6-1. Station Initialization
Sequence
Description
Break State
Figure 6-1 shows the PCM of station C entering the Break
State. During the Break State, Station C sends Quiet Symbols
to the PHY in station B. As a result, Station B stops
transmitting any data or symbols and enters the Break State.
During the Break State, Station B breaks all existing
connections and enters the Quiet Line State.
Quiet Line State
In the Quiet Line State, both stations send Quiet Symbols to
each other. The Quiet Line State means that the stations have
each other’s attention to continue with the initialization
sequence.
Note: If Station B had not entered into the Quiet Line State,
then Station C would have returned to the Break State.
Connect State
During the Connect State each station sends a continuous
stream of Halt Symbols. Halt Symbols synchronize the clocks
in each station.
Halt State
After both stations have had enough time to synchronize
clocks, they enter the Halt State then proceed to the Next State.
Next State/Signal State
The Next State, in conjunction with the Signal State, allows the
two stations to exchange port information (A, B, S, or M type
ports) and port compatibility (S to M, or A to B, etc.).
Note: These two states are also used to exchange information
concerning the Link Confidence Test of the physical
connection and MAC.
Idle Line State
The Idle Line State is used for transitions between the Next
State and the Signal State. When it is ready to receive
information, a station sends continuous Idle Symbols to its
neighbor. If Station C were to enter into the Next State first, it
would send Idle Symbols to station B. Station B would then
enter the Signal State. This particular exchange of information
is accomplished by using Halt Symbols and Master Symbols
(alternating Halt and Quiet Symbol Pairs). The reception of
either Halt Symbols or Master Symbols for more than 30µs
causes the receiving station to enter either the Halt Line State
or the Master Line State. The Halt Line State indicates that the
received bit was a “1,” while the Master Line State indicates a
“0.” The type A port is identified as “00” which means that the
receiving station would record two Master Line States. The
type B port is identified as “01,” which would indicate a
Master Line State followed by a Halt Line State.
6-3
FDDI RING OPERATION
Table 6-1. Station Initialization
Sequence
Description
Link Confidence Test
Link Confidence Test information is also exchanged between
stations. This information indicates the length of the test and
determines if a MAC is involved. If a MAC is not involved in
the test, Idle Symbols are transmitted during the Link
Confidence Test. When there is a MAC involved in the test,
then frames are transmitted as well as procedures to test frame
transmission and reception and token passing and MAC
recovery processes. If the stations pass the Link Confidence
Test then the stations continue to the Join State.
Join State
The Join State ensures that both stations reach the Active State
together. A sequence of line states is used to get the station
from the Join State to the Active State. The detection of Halt
Symbols causes the station to enter the Join State. Detection of
Master Symbols cause the station to enter the Verify State.
finally, the reception of Idle Symbols causes the station to enter
the Active State. Once the PCM has reached the Active State, it
will send a signal to Configuration Management to join the
station to the ring as shown in Figure 6-2.
RING INITIALIZATION
STATION C
STATION B
Link B/C
PHY B
PHY B
MAC
Link C/A
MAC
Link A/B
PHY A
PHY A
PHY A
PHY B
MAC
STATION A
Figure 6-2. The Joint State
6-4
RING INITIALIZATION
After the Station Initialization process is complete and the stations are physically
attached to the ring, as shown in Figure 6-3, the ring must be initialized. The Ring
Initialization procedure determines which station transmits the first token. It also
sets the Operational Token Rotation Time. The Operational Token Rotation Time
determines how long a station must wait before it receives a token. This process
guarantees that each station has a chance to transmit frames onto the ring within a
specific time period.
STATION C
STATION B
Link B/C
PHY B
MAC
PHY B
Link C/A
MAC
Link A/B
PHY A
PHY A
PHY A
PHY B
MAC
STATION A
Figure 6-3. All Stations Physically Attached to the Ring
6-5
FDDI RING OPERATION
Transmitting the First Token
Stations on an FDDI ring bid for the right to issue the first token. Each station has
a preset timing requirement called a Token Rotation Timer (TRT). The TRT
determines how often the token must visit the station. This time is compared with
the timing requirements of the other stations on the ring and the stations bid for
the Target Token Rotation Time (TTRT). The station with the lowest TRT wins the
right to issue the first token.
The bidding procedure begins when each station generates a Claim Frame. This is
a special Frame that includes the sending station’s address and TRT. Each station
compares its own TRT to the TRT on the Claim Frame. If the station’s TRT is lower
than the Claim Frame, then it discards the Frame and issues its own Claim Frame.
If the station’s TRT is higher than the Claim Frame then it repeats the Claim
Frame.
Eventually, the Claim Frame with the lowest TRT remains on the ring and the
station that issued the Claim Frame with the lowest TRT issues the first token. It is
this station that is then allowed to initialize the ring with its winning bid. The
winning bid value becomes the operational Token Rotation Time (T_OPR) for the
ring.
Ring Initialization
Ring Initialization ensures that each station on the ring has the same TRT. This is
done by each station setting their T_OPR to the winning bid value and setting its
Token Rotation Timer (TRT) to the same value. The winning bidder will transmit
the bid around the ring. During this time use of the ring is restricted until all the
stations receive the value of T_OPR. At the completion of this operation a
nonrestricted token is issued.
During the second token rotation each station accumulates the current
synchronous bandwidth from the Token Rotation Timer (how long it took for the
token to circulate the entire ring). Once the synchronous bandwidth has been
calculated and divided between all stations, the remaining bandwidth can be
used for asynchronous transmissions.
6-6
RING INITIALIZATION
Synchronous Transmission
Transmission of normal Protocol Data Units is controlled by a Timed Token
Rotation Protocol. This protocol supports two classes of service, synchronous and
asynchronous. Synchronous service gives each station a guaranteed bandwidth
and response time. Each station can be assured that they see a token once every 2
x T_OPR (two times the winning TTRT bid). It is therefore good for those
applications whose bandwidth and response time limits are predictable in
advance, permitting them to be pre-allocated via station management.
The time budget allowed for transmission is limited to ensure fairness between all
stations. The bandwidth allocation process in Station Management sets this
maximum transmission time and is a percentage of the total time allocated for
synchronous transmission.
Asynchronous Transmission
This type of service is used for those applications whose bandwidth requirements
are less predictable (e.g. “bursty” or potentially unlimited) or whose response
time requirements are less critical.
Asynchronous bandwidth is instantaneously allocated from a pool of remaining
bandwidth. Transmission time is not guaranteed but is related to the activity of all
the stations on the ring. Once transmission has started, the time a station is
allowed to transmit is limited to the value of a timer called the Token Holding
Timer (THT). There is a two tier allocation mechanism for controlling
Asynchronous bandwidth which is achieved by using two classes of tokens
Restricted an Nonrestricted Tokens.
Token Priorities
Different priority levels can be set within the asynchronous transmissions
depending on the value of TRT when the token is received.
Within each station, the MAC transmitter maintains a Token-Rotation Timer
(TRT) to control the ring scheduling. TRT is reset each time a token arrives. A
token arriving at a station before the TRT reaches the T_OPR value is called an
“Early Token.” An early token can be used for either Synchronous or
Asynchronous transmission.
A token arriving after TRT reaches the value of T_OPR is known as a “Late
Token” and this type of token can only be used for Synchronous transmission.
Different mechanisms may be used to limit the length of a station’s transmission
whether it be Synchronous or Asynchronous, but in no case should the station
hold the Token longer than the negotiated Token Holding Timer (THT).
6-7
FDDI RING OPERATION
Restricted/Non-Restricted Token Mode
Asynchronous bandwidth is controlled by a two tier allocation mechanism,
enforced by two classes of tokens. Normal operation is achieved using a
Non-Restricted Token being issued by a station.
Restricted Token Mode may be entered when a station wishes to initiate an
extended dialogue requiring substantially all of the unallocated ring bandwidth
(e.g., an extended burst data transfer from a high speed device). If a station
receives a non-restricted token that is very early it can transmit the initial part of
an extended dialogue to the assigned destination station and it will then issue a
restricted token.
The destination station receives the initial dialogue and will enter restricted token
mode. When the restricted token arrives at this station now, it can reply and then
issue a restricted token. Both stations can now exchange frames and restricted
tokens for the duration of the dialogue (this could typically be many times the
negotiated T_OPR value).
Restricted Token mode is terminated when a station captures a restricted token,
transmits its final dialogue frame(s) and then issues a non-restricted token. Since
fairness is the name of the game within the FDDI protocol, there is no need to use
a Token Holding Timer for the extended dialogue. However the Station
Management entities negotiate and monitor a maximum restricted Token mode
time. If restricted Token mode operation exceeds this time, SMT should abort the
extended dialogue.
The advantage of using Restricted token mode is that it is fair to all stations on the
ring. Each station on the ring that requires an extended dialogue has an
opportunity to do so.
Ring Timing and Latency
Transmission has to be guaranteed under synchronous operation and the protocol
ensures that the right to transmit will occur by twice the negotiated T_OPR value.
To achieve this, the ring latency (the propagation delay of transmitted data
around the ring) and the total transmission time by the stations must be limited.
Maximum Ring Latency
The maximum ring latency is the maximum delay introduced into the ring by
each of the stations and the cable that makes up the physical ring. This delay can
be calculated using the maximum allowable rules that are stated within the FDDI
specification.
The maximum ring distance (allowing for a ring in the wrapped state) is 200 km.
and the maximum number of connections allowed on this ring is 1000. The
maximum ring latency equals 1.617 ms (this is the default in the FDDI
specification). The formula for Maximum Ring Latency is:
Maximum Ring Latency = Total Station Delay + Total Fiber Delay.
6-8
BASIC RING OPERATION
Total Transmission Time
The total allowable transmission time is the negotiated value of T_OPR minus the
maximum ring latency. The formula for Total Transmission Time is:
Total Transmission Time = T_OPR - Maximum Ring Latency.
Token Rotation Timer
Each station has a timer called the Token Rotation Timer (TRT) that is used to
control ring scheduling during normal operation and to detect and recover from
serious ring error situations. TRT is initialized with different values during
different phases of ring operation. Whenever TRT expires, it is reinitialized to the
current value of T_OPR.
Token Hold Timer
Each station has a timer called the Token Hold Timer (THT) that controls how
long the station is allowed to transmit asynchronous frames. THT is initialized
with the current value of TRT when a token is captured.
BASIC RING OPERATION
Once the stations have completed all their self tests the Station Management
works with the Physical Layer Protocol (PHY) and the configuration logic to
connect adjacent PHY entities and perform a handshake by the transmission of
line state symbols as previously discussed. If the links are good, then the “mini
rings” are joined until a fully configured ring is achieved.
Once the ring is established, stations take part in the Claim Process as previously
discussed. As a result of this process, the station bidding the lowest the Target
Token Rotation Time initializes the ring by issuing a non-restricted token. Once
this token has been released other station on the ring can begin the transmission
of frames.
6-9
FDDI RING OPERATION
The Beacon Process
The Beacon process is used to recover from serious ring faults. These faults
include a failed Claim Process, a broken ring, re-configured ring, or the joining of
two logical rings into one. The purpose of the Beacon Process is to signal to all
other stations that a significant logical break in the ring has occurred and to
provide diagnostics or other assistance to recover the ring using SMT.
Upon entering the Beacon state a station continuously transmits MAC Beacon
frames to it’s downstream neighbor. All stations on the ring repeat Beacon Frames
so there comes a time when the only station beaconing is the station downstream
from the logical break.
While the beaconing process is operating the Connection Management portion of
SMT will have noticed a loss of link at the PHY level and hence will have no line
activity. Connection Management informs Station Management and then sends
control symbols around the ring to its upstream neighbor (it already knows this
address as the upstream neighbor because all stations transmit special MAC
frames called Neighbor Information Frames (NIF) to their downstream MACs on
a periodic basis).
The station on the other side of the break receives the beacon frames and under
control of SMT removes from the ring to perform local tests. If these tests pass, it
re-enters the ring. The beaconing station still has not seen its own frames
returned, so it removes itself from the ring and perform some basic tests.
With the satisfactory completion of these tests, Station Management will force the
relevant connections into the Wrap state. Both the primary and the secondary
rings will now be used to recover the ring and the beaconing station will receive
its beacon frames back. Once the beaconing station receives the transmitted
frames, it knows the ring is now re-configured and the initialization process can
begin.
Frame Transmission
When an FDDI node needs to transmit a frame onto the ring it must first capture
the Token as it enters the node, The node transmits the frame(s) onto the ring and
then transmits the Token. The frame will circulate around the ring, being repeated
by each node, until it reaches the destination node, This node recognizes the
frames’s destination address and copies the frame into its receive buffer; and then
repeats the frames back onto the ring, The frame will continue to circulate the ring
and then is stripped from the ring by the originating node.
6-10
BASIC RING OPERATION
Ring Fault Recovery
An FDDI network consists of two distinct separate rings., the primary ring and
the secondary ring. Under normal conditions data frames travel on the primary
ring and the secondary ring is used as a back-up path. If a fiber is cut between two
Dual Attached nodes its upstream and downstream neighbor will wrap. When a
node wraps a port it internally connects the primary ring to the secondary ring.
This maintains a data path for frame transmission by effectively creating one
contiguous enveloped ring. Similarly, if a Dual Attached node is de-powered or
de-inserted, the upstream and downstream neighbors will wrap.
Since FDDI networks employ a ring topology, the entire network is vulnerable to
the frailties of each ring segment and failures of the individual stations. The ring
of trees topology reduces the risk of a single node bringing the entire network
down. To further reduce this vulnerability, a redundant data path is provided in
the main ring trunk cabling. In theory, the ring topology requires media that is
capable of only one-way traffic to achieve the circular flow of data. In practice, an
FDDI ring uses media that provides two fiber optic ring paths, a primary ring and
a secondary ring. The secondary ring is used to restore the continuity of the ring
in the event of a failed node or trunk segment (broken trunk cable). Figure 6-4
shows how the open ends of the primary ring are wrapped into the secondary
ring, restoring continuity through the creation of a new ring.
RING WRAP
DUAL ATTACHED
CONCENTRATOR 3
CABLE FAILURE
PRIMARY RING
A
B
DUAL ATTACHED
CONCENTRATOR 4
SECONDARY RING
A
B
A
DUAL ATTACHED
CONCENTRATOR 2
FDDI DUAL
COUNTER-ROTATING RING
NETWORK
B
B
A
DUAL ATTACHED
CONCENTRATOR 1
Figure 6-4. Wrapping a Broken Ring
6-11
FDDI RING OPERATION
OPTICAL BYPASS SWITCH
In a network design where the wrap function is undesirable, an Optical Bypass
Switch (OBS) is used to maintain the integrity of the data flow through the
network while not combining the primary and secondary rings. When a node is
de-powered or is malfunctioning, the OBS is activated and diverts the frames
through the switch instead of to the node, This eliminates the need for upstream
and downstream nodes to wrap. However, some rules need to be followed to stay
within the maximum distance between nodes, The maximum distance between
nodes, using multimode fiber, is 2 km. When designing an OBS into your network
you must determine the worst case distance between nodes and calculate in the
optical loss introduced by the OBS itself, usually 2.5 db. An OBS is an excellent
option for protection against node failures when these simple configuration rules
are applied.
Figure 6-5 shows the data paths through the switch in both the bypass and
operational (non-bypassed) states.
BYPASS STATE
OPERATIONAL STATE
Station Power Off
Station Power On
Station
Station
FDDI Dual
Optical Bypass
Switch
FDDI Dual
Optical Bypass
Switch
FDDI
Ring
FDDI
Ring
Figure 6-5. Optical Bypass Switch
6-12
Chapter 7
BRIDGING WITH THE FDMMIM
This chapter explains how Cabletron’s FDMMIM performs translational bridging
from Ethernet to FDDI and back to Ethernet.
ETHERNET FRAME TYPES
There are four basic Ethernet frame types:
•
802.3 "raw"
•
Ethernet II (DIX)
•
Ethernet 802.2
•
Ethernet SNAP
Figure 7-1 shows each of the Ethernet frame types. The Ethernet 802.2 and
Ethernet SNAP frames are extensions of the 802.3 "raw" frame format, while the
Ethernet II frame is formatted slightly different. The following sections describe
each frame type.
7-1
BRIDGING WITH THE FDMMIM
Ethernet II Frame
Preamble
Destination
Address
Source
Address
Type
Data
FCS
8 bytes
6 bytes
6 bytes
2 bytes
46 - 1500 bytes
4 bytes
802.3 "Raw" Frame
Preamble
Start Frame
Delimiter
Destination
Address
Source
Address
Length
Data
FCS
7 bytes
1 byte
6 bytes
6 bytes
2 bytes
46 - 1500 bytes
4 bytes
Ethernet 802.2 Frame
Preamble
Start Frame
Delimiter
Destination
Address
Source
Address
Length
DSAP
SSAP
Control
Data
FCS
7 bytes
1 byte
6 bytes
6 bytes
2 bytes
1 byte
1 byte
1 byte
43 - 1497 bytes
4 bytes
Ethernet SNAP Frame
Preamble
Start Frame
Delimiter
Destination
Address
Source
Address
Length
DSAP
SSAP
Control Protocol
Identifier
7 bytes
1 byte
6 bytes
6 bytes
2 bytes
1 byte
1 byte
1 byte
5 bytes
Data
FCS
37 - 1492 bytes
4 bytes
Figure 7-1. Ethernet Frame Type
The Length Field value does not include the length of the SFD and or the Preamble
NOTE
7-2
ETHERNET FRAME TYPES
Ethernet II Frame Type
The Ethernet II or (DIX) frame format was developed by DEC, Intel Corporation,
and Xerox Corporation. Table 7-1 describes each of the Ethernet II Frame fields.
Table 7-1. Ethernet II Frame Type
Field Name
Field Size
Field Definition
Preamble
8 bytes
Signals beginning of the packet.
Destination Address
6 bytes
Address of the destination of the packet.
Source Address
6 bytes
Address of the packets origin.
Type Field
2 bytes
Specifies the upper layer protocol used.
Data
46 - 1500 bytes
Contains the data to be transferred.
Frame Check
Sequence
4 bytes
Determine the integrity of the packet.
The Ethernet II frame uses a frame Type Field in place of the Length Field used in
the 802.3 "raw" format. The Type Field is a 2 byte field that specifies the higher
layer protocol used in the Data Field (XNS, DecNet, TCP/IP etc.) often called the
"Ethertype". The total length of the packet may range from 64 to 1518 bytes not
including the Preamble.
Ethernet “Raw” Frame Type
The total length of the packet may range from 64 to 1518 bytes, not including the
Preamble and Start Frame Delimiter. Table 7-2 describes each of the Ethernet
“Raw” Frame Type fields:
Table 7-2. Ethernet “Raw” Frame Type
Field Name
Field Size
Field Definition
Preamble
7 bytes
Signals beginning of the packet.
Start Frame Delimiter
1 byte
Signals start of data.
Destination Address
6 bytes
Address of the destination of the packet.
Source Address
6 bytes
Address of the packets origin.
Length Field
2 bytes
Specifies the length of the data field.
Data
46 - 1500 bytes
Contains the data to be transferred.
Frame Check
Sequence
4 bytes
Used to determine integrity of the packet.
7-3
BRIDGING WITH THE FDMMIM
Ethernet 802.2 Frame Type
The Ethernet 802.2 frame format builds upon the 802.3 "raw" frame structure.
Table 7-3 describes each of the Ethernet 802.2 frame fields.
Table 7-3. Ethernet 802.2 Frame Type
Field Name
Field Size
Field Definition
Preamble
7 bytes
Signals beginning of the packet.
Start Frame Delimiter
1 byte
Signals start of data.
Destination Address
6 bytes
Address of the destination of the packet.
Source Address
6 bytes
Address of the packets origin.
Length Field
2 bytes
Length of the Data plus LLC fields.
Destination Service
Access Point (DSAP)
1 byte
First byte of 2 byte value indicating the
packet’s upper layer protocol destination.
(Source Service
Access Point) SSAP
1 byte
Second byte of 2 byte value indicating the
packet’s upper layer protocol destination.
Control
1 byte
Indicates the type of LLC frame.
Data
43 - 1497
bytes
Contains the data to be transferred.
Frame Check
Sequence
4 bytes
Used to determine integrity of the packet.
The Ethernet 802.2 frame adds a Logical Link Control (LLC) header directly after
the Length Field. This LLC header or 802.2 header consists of a 1 byte Destination
Service Access Point (DSAP) field, a 1 byte Source Service Access Point (SSAP)
field, and a 1 byte Control field. Protocols are assigned hexadecimal values which
are displayed in the DSAP and SSAP fields of a packet. For example, packets
containing Novell's IPX/SPX will display values of E0 in the DSAP and SSAP
fields. The total length of the packet may range from 64 to 1518 bytes not
including the Preamble and SFD.
7-4
ETHERNET FRAME TYPES
Ethernet SNAP Frame Type
The Ethernet Sub-Network Access Protocol (SNAP) frame type is an extension of
the Ethernet 802.2 frame structure. The Ethernet SNAP frame adds a 5 byte
Protocol Identification Field immediately following the LLC header. The Protocol
Identification Field is made up of a 3 byte Protocol ID or Organizational Code
Field followed by a 2 byte Type Field "Ethertype". In accordance with RFC 1042,
SNAP frames are transmitted with the DSAP and SSAP fields set to AA (hex) and
the Control Field set to 03 (hex). The SNAP frame uses the Protocol Identifier to
determine which upper layer protocol the frame is intended for. Table 7-4
describes each of the Ethernet SNAP frame fields.
Table 7-4. Ethernet SNAP Frame Type
Field Name
Field Size
Field Definition
Preamble
7 bytes
Signals beginning of the packet.
Start Frame Delimiter
1 byte
Signals start of data.
Destination Address
6 bytes
Address of the destination of the packet.
Source Address
6 bytes
Address of the packets origin.
Length Field
2 bytes
Length of the Data plus LLC fields.
Destination Service
Access Point (DSAP)
1 byte
Set to AA (hex) and 10101010 (binary)
Source Service Access
Point. (SSAP)
1 byte
Set to 03 (hex) and 00000011 (binary)
Control Field
1 byte
Indicates the type of LLC frame.
Protocol Identifier
5 bytes
Specifies the upper layer protocol.
Data
43 - 1497
bytes
Contains the data to be transferred.
Frame Check
Sequence
4 bytes
Determines integrity of the packet.
The total length of the packet must be a minimum of 64 bytes in length, with a
maximum size limit of 1518 bytes not including the Preamble and SFD.
7-5
BRIDGING WITH THE FDMMIM
FDDI FRAME TYPES
Figure 7-2 shows an FDDI Token Frame and an FDDI Data Frame:
TOKEN
Preamble
≥ 16 Symbols
Starting
Delimiter
2 Symbols
Frame
Control
2 Symbols
Preamble
≥ 16 Symbols
Starting
Delimiter
2 Symbols
Frame Check
Sequence Coverage
T T
J K
FRAME
Ending
Delimiter
2 Symbols
Frame
Control Destination Address
2 Symbols 4 or 12 Symbols
Source Address
4 or 12 Symbols
T
Information
≥ 0 Symbol Pairs
Frame Check Ending
Delimiter
Sequence
8 Symbols 1 Symbol
Frame Status
≥ 3 Symbols
Maximum - 9000 symbols
Figure 7-2. FDDI Frames
Table 7-5. FDDI Frame
Field Name
Field Definition
Preamble
16 + symbols
Signals the start of a valid frame.
Start Delimiter
2 Symbols
Signals that FC is next field.
Frame Control
2 Symbols
Identifies the type of frame (MAC, LLC, etc.).
Destination Address
4 or 12 Symbols
Address of the destination of the packet.
Source Address
4 or 12 Symbols
Address of the packets origin.
Information (Data)
≤ 8956 Symbols
Contains the data to be transferred.
FCS (Frame Check
Sequence)
8 Symbols
Used to determine integrity of the packet.
ED (Ending
Delimiter)
1 Symbol
Signals the end of the frame.
FS (Frame Status)
3 Symbols
Indicates the status of the frame.
NOTE
7-6
Field Size
FDDI uses a 5 bit symbol scheme. The PHY handles the encoding and decoding of the
four bit to five bit symbols.
FDDI FRAME TYPES
FDDI 802.2 Frame Type
There are two basic FDDI frame types that are used for the transmission of data:
•
802.2
•
FDDI SNAP
The FDDI 802.2 frame type contains the same 802.2 (LLC) header set up as the
Ethernet 802.2 frame type. The differences between the two frames are due to
their technological differences. These differences can be seen when comparing the
802.3 header with the FDDI header. The FDDI frames have a Frame Control field
and no length field, while the Ethernet frames do not have a Frame Control field,
but do have a Length/Type field. Notice that the 802.2 or LLC headers are
identical in both the Ethernet 802.2 and FDDI 802.2 frame formats.
Table 7-6. FDDI 802.2 Frame Type
Field Name
Field Size
Field Definition
Frame Control
2 bytes
Signals beginning of the packet.
Destination Address
6 bytes
Address of the destination of the packet.
Source Address
6 bytes
Address of the packets origin.
DSAP
1 bytes
Destination service access point.
SSAP
1 bytes
Source Service Access Point.
Control
1 bytes
Indicates the type of LLC frame.
Data
43 - 4478
bytes
Contains the data to be transferred.
Frame Check
Sequence
4 bytes
Used to determine integrity of the packet.
7-7
BRIDGING WITH THE FDMMIM
FDDI SNAP Frame Type
The FDDI SNAP frame builds upon the 802.2 (LLC) layer of the FDDI 802.2 frame,
just as the Ethernet SNAP frame builds upon the LLC layer of the Ethernet 802.2
frame. The SNAP header consists of the 802.2 header plus a Protocol
Identification Field. The FDDI and Ethernet SNAP frame differ in the same way
the FDDI and Ethernet 802.2 frame types differ. The Ethernet SNAP frame does
not have a Frame Control Field, but does have a Length Field. The FDDI SNAP
frame does have a Frame Control Field, but does not contain a Length Field.
Notice that the SNAP headers of the two frames are identical.
There is also a third frame type discussed in this document called a FDDI “MAC”
frame. This frame type is not a legal data frame on FDDI, but is used as a work
around which is discussed latter in this document.
Table 7-7. FDDI SNAP Frame Type
Field Name
7-8
Field Size
Field Definition
Frame Control
2 bytes
Identifies the type of frame (MAC, LLC, etc.)
Destination Address
6 bytes
Address of the destination of the packet.
Source Address
6 bytes
Address of the packets origin.
DSAP
1 byte
Destination Service Access Point.
SSAP
1 byte
Source Service Access Point.
Control
1 byte
Indicates the type of LLC frame.
Protocol Identifier
5 bytes
Specifies the upper layer protocol.
Data
up to 4478
bytes
Contains the data to be transferred.
Frame Check
Sequence
4 bytes
Used to determine integrity of the packet.
ETHERNET TO FDDI BRIDGING
ETHERNET TO FDDI BRIDGING
When bridging a frame from Ethernet to FDDI the FDMMIM has to deal with 4
different types of Ethernet frames, and translate them into one of two FDDI frame
formats*. The FDMMIM's first task is to determine what type of frame it is
receiving on Ethernet. To do this the FDMMIM goes directly to the 2 byte field
immediately following the Source Address of the packet. If the decimal value of
this two byte field is greater than 1500 bytes, the FDMMIM knows that the frame
is an Ethernet II frame. How can the FDMMIM draw this conclusion? The
FDMMIM can do so because the 2 byte field immediately following the Source
Address in an Ethernet II frame is the Type Field. The Hex values assigned to
protocols that are inserted into the Type Field are greater than 1500 decimal. The
maximum decimal value of a frames Length Field is 1500, the maximum size of
the data portion of an Ethernet packet. When the FDMMIM reads a value greater
than 1500 decimal in this field of a packet, it knows that it is a Type Field and not
a Length Field. Since there is a Type Field in place of a Length Field the FDMMIM
determines that the frame is in the Ethernet II format.
When the value of the 2 byte field immediately following the Source Address of a
packet is equal to or less than 1500 decimal, then the FDMMIM knows that the
frame falls under the 802.3 suite. Meaning that the frame is either an 802.3 “raw”,
802.2, or a SNAP Ethernet frame. When the FDMMIM detects one of these frame
types on Ethernet it simply removes the Length Field, puts the frame into FDDI
format, then transmits the frame onto the FDDI ring.
The following sections describe how the FDMMIM bridges a frame from Ethernet
to FDDI.
NOTE
*A third frame type is used when not enough information is provided by the Ethernet
frame to construct an 802.2 or SNAP frame on FDDI.
7-9
BRIDGING WITH THE FDMMIM
Ethernet II to FDDI SNAP Frame Bridging
When an FDMMIM bridges an Ethernet II frame onto FDDI it will insert a SNAP
header into the frame and format it as an FDDI SNAP frame. The following is an
illustration of this process. In this particular example there is a workstation
located on an Ethernet segment (00001D09535D) which is transferring a file from
a Novell server (00001D08A333) which is located on another Ethernet segment.
The two Ethernet segments are connected via an FDDI backbone using
FDMMIMs. When the FDMMIM receives the frame on its Ethernet port it will
view the 2 byte field immediately following the Source Address. This field
contains a value of 8137 (hex) [Novell's assigned #] which is 33079 decimal. Since
33079 is greater the 1500 the FDMMIM knows the incoming frame is an Ethernet
II frame. Since it is an Ethernet II frame type the FDMMIM will add a SNAP
header and put the frame into FDDI format. Notice that a Frame Control Field
has been added. The FDMMIM will also recalculate the Frame Check Sequence
and insert the Start of Frame Sequence and End of Frame Sequence as show on
page 7. Also remember that the DSAP and SSAP are set to AA (hex) and the
Control set to 03 (hex) when a SNAP header is in place. Novell’s protocol
identifier is equal to 00008137 (hex).
802.3 “Raw” Frame to FDDI MAC Frame Bridging
Now lets consider the FDMMIM receiving an 802.3 "raw" Frame on Ethernet.
Novell's IPX is the only protocol that uses the 802.3 "raw" frame type. The
FDMMIM will look at the 2 byte field immediately following the Source Address.
The decimal value of 40 (hex) is equal to 64 which is less than 1500. Since this
value is less than 1500 the FDMMIM will remove the Length Field and put the
frame into FDDI format. It does so by adding the Start of Frame Sequence, Frame
Control Field, and End of Frame Sequence. This causes somewhat of a problem
when the frame is passed onto the FDDI ring. The end result is an FDDI frame
that has neither an 802.2 or SNAP header. This problem is a result of the 802.3
“raw” frame not having a Type Field, 802.2, or SNAP header. However, other
FDMMIMs on the ring will recognize this frame along with many other vendor's
Ethernet to FDDI bridges. This frame is recognized as a special case and will be
forwarded onto Ethernet segments as an 802.3 "raw" frame.
802.2 Frame To FDDI 802.2 Frame Bridging
Now lets take a look at what happens with an Ethernet 802.2 Frame Type. Again
the FDMMIM will go to the 2 byte field immediately following the Source
Address. It contains a hex value of 40. The hex value 40 is equal to 64 decimal,
which is less that 1500. Since the FDMMIM knows that the frame is not an
Ethernet II frame it will simply take out the Length Field, add the Frame Control,
Start of Frame Sequence, and End of Frame Sequence and transmit the frame onto
the FDDI ring. The 802.2 header will be kept intact. As mentioned on page 5 of
this document the DSAP and SSAP fields in an 802.2 frame are used to indicate
what upper layer protocol the frame is destined for. Notice the DSAP and SSAP
values are E0 (hex), this is Novell's assigned identification number.
7-10
FDDI TO ETHERNET BRIDGING
Ethernet SNAP Frame to FDDI SNAP Frame Bridging
Finally, we have the Ethernet SNAP frame type. Again the FDMMIM will go to
the 2 byte field immediately following the Source Address. It contains a hex value
of 40. The hex value 40 is equal to 64 decimal, which is less than 1500. Since the
FDMMIM knows that the frame is not an Ethernet II frame it will simply take out
the Length Field, add the Frame Control, Start of Frame Sequence, and End of
Frame Sequence. Once this is done it will transmit the frame onto the FDDI ring.
The SNAP header is kept intact. The end result is an FDDI SNAP frame type.
Notice that the DSAP and SSAP fields contain AA (hex) values. This is a Novell
frame therefore the organizational code is all zeros (first 4 bytes), and the Ether
type is 8137 (last 2 bytes) in the Protocol Identifier.
FDDI TO ETHERNET BRIDGING
The following sections examine what the FDMMIM does when bridging frames
from FDDI to Ethernet. The biggest concern that we have when bridging a frame
from FDDI to Ethernet is that we must maintain consistency with frame types.
When a frame originates as an 802.3 "raw" frame on one Ethernet segment and
gets bridged onto FDDI, the FDMMIM must bridge the frame back to Ethernet as
an 802.3 "raw" frame. The same must happen when bridging Ethernet II, Ethernet
802.2, and Ethernet SNAP frames.
Since the FDMMIM is a bridge and not a router, guidelines had to be developed to
help achieve frame type consistency. There are four possible cases that the
FDMMIM has to handle. Case 1: a frame originates on Ethernet as an 802.3 "raw"
frame and gets bridged onto FDDI as a FDDI "MAC frame. Case 2: a frame
originates on Ethernet as an Ethernet 802.2 frame and gets bridged onto FDDI as a
FDDI 802.2 frame. Case 3: a frame originates on Ethernet as an Ethernet II frame
type and gets bridged onto FDDI as a FDDI SNAP frame. Case 4: a frame
originates on Ethernet as an Ethernet SNAP frame and gets bridged onto FDDI as
a FDDI SNAP frame. The following pages describe the guidelines that the
FDMMIM follows, and also some problems that can occur and why.
7-11
BRIDGING WITH THE FDMMIM
Case 1
A frame originated as an 802.3 "raw" frame on Ethernet and has been bridged
onto FDDI. The FDMMIM that is going to forward this frame onto an Ethernet
segment is responsible for forwarding the frame as an 802.3 "raw" frame. In reality
the format of the frame on FDDI is not a legal FDDI data frame. Because Novell is
using the 802.3 "raw" format, Ethernet to FDDI bridge vendors had to develop a
way to pass these frames onto FDDI. Remember that the only legal data frame
types on FDDI are FDDI 802.2 and FDDI SNAP. The problem exists because there
is no 802.2 header, SNAP header, or Type Field to provide protocol information.
The protocol information is needed since the only legal data frames allowed on
FDDI are FDDI 802.2 and FDDI SNAP. To get around this, the FDMMIM will put
the frame on the ring as an FDDI "MAC" frame. The FDMMIM will recognize this
frame as a Novell "raw" frame and forward it properly. The FDMMIM will
recalculate the length of the packet, insert the Length Field, add a Preamble and
Start Frame Delimiter, add a new Frame Check Sequence, and then transmit the
frame onto Ethernet. Most Ethernet to FDDI bridge vendors use the same method
when dealing with Novell's 802.3 "raw" frames.
Case 2
A frame originates on an Ethernet segment as an Ethernet 802.2 frame and gets
bridged to FDDI as a FDDI 802.2 frame. The FDMMIM responsible for bridging
this frame to Ethernet will look at the 2 bytes directly following the Source
Address of the frame. If it does not detect AAAA (hex) which would be the DSAP
and SSAP, it knows that this frame does not contain a SNAP header. Since the
frame does not contain a SNAP header the FDMMIM will recalculate the length of
the packet and insert a Length Field, remove the Frame Control Field, add a
Preamble and Starting Delimiter, and then insert a new Frame Check Sequence.
The 802.2 header is kept intact and is passed on to the Ethernet frame. The end
result is an Ethernet 802.2 frame type.
Case 3
In this case a frame originated on an Ethernet segment in the Ethernet II format
and was bridged to FDDI as a FDDI SNAP frame. The FDMMIM responsible for
bridging this frame to Ethernet will look at the 2 bytes directly following the
Source Address of the frame. These two bytes contain AAAA (hex), which are
the DSAP and SSAP fields. This indicates to the FDMMIM that the frame contains
a SNAP header. FDMMIMs forward FDDI SNAP frames as Ethernet II frames,
except for one exception which will be discussed in case 4. The last 2 bytes of the
Protocol Identifier make up the "Ethertype" Field, which denotes the protocol
being used. The "Ethertype" Field of a SNAP frame contains the same hex values
that would go into the Type Field of an Ethernet II frame. Therefore, the
FDMMIM strips off all of the SNAP header except for the last 2 bytes. The
FDMMIM will also remove the Frame Control field, add a Preamble and a new
Frame Check Sequence, then transmit the frame onto Ethernet. The end result is
an Ethernet II frame type.
7-12
FDDI TO ETHERNET BRIDGING
Case 4
This is a special case that is handled by the FDMMIM. The frame originates as an
Ethernet SNAP frame and is bridged to FDDI. Case 3 tells us that when the
FDMMIM bridges a FDDI SNAP frame to Ethernet it translates the frame into an
Ethernet II frame. There is an exception to that rule. When the FDMMIM detects
an FDDI SNAP frame with 809b hex (AppleTalk over Ethernet) or 80f3 hex
(AppleTalk ARP) in the last 2 bytes of the Protocol Identifier Field, it bridges the
frame to Ethernet as an Ethernet SNAP frame. Why does the FDMMIM handle
these frames differently? The answer is because AppleTalk and AppleTalk ARP
over Ethernet use the Ethernet SNAP frame format. These two protocols are the
only ones that predominately use the Ethernet SNAP frame type. Most protocols
in today's Ethernet networks use either the 802.2 or the Ethernet II frame type,
except for Novell's 802.3 "raw" frames. Since FDDI supports the 802.2 frame
format it was easy to maintain a consistent frame format when bridging the 802.2
frame types. However, when bridging Ethernet II frames the conversion is made
to FDDI SNAP and then back to Ethernet II. This method was incorporated
because of the popularity of the Ethernet II frame format. Because the FDDI
SNAP was being converted back to Ethernet as Ethernet II frames, a special case
had to be developed to handle the AppleTalk and AppleTalk ARP protocols over
Ethernet. These protocols needed the Ethernet SNAP frame type. The Ethernet
SNAP frame is bridged to FDDI as a FDDI SNAP frame, then it would normally
be bridged back to Ethernet as an Ethernet II frame. As you can see this would
create some problems. Therefore, when the FDMMIM receives a FDDI SNAP
frame containing the AppleTalk over Ethernet (809b) or AppleTalk ARP (80f3) as a
protocol type, it will bridge the frame keeping the SNAP header intact resulting in
an Ethernet SNAP frame.
The FDMMIM receives a frame with a SNAP header. The last 2 bytes of the
Protocol Identifier is either 809b or 80f3. The FDMMIM realizes that this is a
special case and will strip the Frame Control, add a Preamble and Starting
Delimiter, recalculate and insert a Length Field, keep the SNAP header intact, add
a new Frame Check Sequence, and then transmits the frame onto Ethernet. The
end result is an Ethernet SNAP frame.
Table 7-8 is shows how the FDMMIM translates different frame types. The table
portrays a frame being generated on Ethernet getting bridged to FDDI, and then
bridged back to an Ethernet segment.
Table 7-8. Ethernet/FDDI Bridging
Ethernet Segment 1
FDDI Ring
Ethernet Segment 2
Ethernet II
FDDI SNAP
Ethernet II
802.3 “Raw”
FDDI MAC
802.3 “Raw”
Ethernet 802.2
FDDI 802.2
Ethernet 802.2
Ethernet SNAP
(Appletalk/Appletalk ARP)
Ethernet SNAP
(Appletalk/Appletalk ARP)
Ethernet SNAP
(Appletalk/Appletalk ARP)
7-13
BRIDGING WITH THE FDMMIM
The following combinations of frame types may cause problems:
7-14
•
If a station generates an Ethernet SNAP frame when using any protocol other
than the AppleTalk or AppleTalk ARP protocols there will be problems with
frame type consistency. The Ethernet SNAP frame will get bridged to FDDI as
a FDDI SNAP frame, however it will get bridged back to Ethernet as an
Ethernet II frame. This is because the last 2 bytes of the Protocol Identifier of
the FDDI SNAP frame will not be 809b or 80f3.
•
There may be a problem when existing Novell networks begin placing stations
or servers on the FDDI ring. Versions of Novell's Netware lower than 4.00 have
802.3 "raw" as a default frame type. Stations that are put onto the FDDI ring
will not be able to use this frame type. The stations located on FDDI will have
to use either the FDDI 802.2 frame type or the FDDI SNAP frame type. This
will cause problems with stations on FDDI trying to communicate with
stations on Ethernet using the 802.3 "raw" frame type, and visa versa. To avoid
this problem Novell recommends that all stations, regardless of there topology
be re configured to use the 802.2 frame format. Netware 4.00 now defaults to
the 802.2 frame format, as well as the 802.3 "raw".
•
There also may be problems when FDMMIMs are on the same ring as routers,
and the routers are configured to use only one frame type, such as FDDI SNAP.
The routers will have to be configured to support the correct frame types on
Ethernet and FDDI as well as the end stations. If bridging is enabled on the
routers there should not be any problems, since most vendors use the same
type of translational bridging as Cabletron. Should the router use the same
type of translational bridging as the FDMMIM we must still be aware of
problems 1 and 2.
Appendix A
ANSI STANDARDS FOR FDDI
This chapter describes the American National Standards Institute (ANSI)
standards for FDDI. ANSI is the governing body of FDDI standards and all
devices on an FDDI ring must comply with these standards. The ANSI standards
committee defines the following entities:
•
Station Management (SMT) - ANSI X3T9.5
•
Media Access Control (MAC) - ANSI X3.139
•
Physical Layer Protocol (PHY) - ANSI X3.148
•
Multimode Fiber Physical Layer Medium Dependant (PMD) - ANSI X3.166
•
Single Mode Fiber Physical Layer Medium Dependent (SMF-PMD) - ANSI
X3.184
•
Twisted Pair Physical Layer Medium Dependent (TP-PMD) - ANSI
X3T9.5/94-044
Each entity performs tasks which are essential to the operation of the FDDI
network including media access, token passing, and frame generation. The
entities defined by ANSI perform many of the functions required in the
International Standards Organization (ISO) Open Systems Interconnection (OSI)
network model. The following sections describe the OSI network model as well as
each of the entities defined by the ANSI FDDI standard
A-1
ANSI STANDARDS FOR FDDI
THE OSI NETWORK MODEL
The OSI network model defines standards for communication between computer
equipment and networks.The FDDI entities defined by ANSI perform many of the
functions required in layer 1 (Physical) and Layer 2 (Data Link) of the OSI
network model. Figure A-2 shows the relationship between the OSI Model and
the ANSI FDDI entities.
OSI MODEL
Layer 7 - Application
ANSI STANDARDS
Layer 6 - Presenation
Layer 5 - Session
SMT
Station Management
-Fault Isolation and
Recovery
-Station Configuration
-Scheduling Procedures
Layer 4 - Transport
Layer 3 - Network
Layer 2 - Data Link
MAC
Media Access Control
Layer 1 - Physical
-Medium Addressing
-Data Checking
-Data Framing
SMT
PHY
Physical Layer Protocol
-Symbol Coding/Decoding
-Symbol Framing
-Clock Rate
SMT
PMD or SMF-PMD
Physical Layer Medium Dependent
-Power Levels
-Transmitter
-Receiver
-Optical Interface
-Connector Types
SMT
Figure A-1. ANSI FDDI Standards and the OSI Network Model
A-2
STATION MANAGEMENT (SMT)
STATION MANAGEMENT (SMT)
SMT is the management entity. It communicates with the MAC, PHY, and PMD
entities to ensure proper station and ring operation. SMT communicates with the
SMT of each station on the FDDI network to ensure proper ring operation. The
ANSI X3T9.5 specifies three distinct SMT functions:
•
•
•
SMT Frame Services
Connection Management
Ring Management
Each FDDI Station may have several instances of MAC, PMD, or PHY but may
have only one instance of SMT. The following sections define each SMT function.
SMT Frame Services
SMT Frame Services provide Frame Based Management Protocols to
communicate with network management and the SMT of each station on an FDDI
network. SMT Frame Services defines the following protocols:
SMT Management Information Base (MIB)
Simple Network Monitoring Protocol (SNMP) is a management protocol that
allows system managers to control and monitor a network using Management
Information Base (MIB) variables. The SMT MIB holds all the read/write data for
SMT.
A-3
ANSI STANDARDS FOR FDDI
SMT Frame Based Management Protocols
SMT Frame Based Management Protocols allow FDDI stations to communicate
with the SMT of other FDDI stations on a ring. It gathers network statistics as well
as detects, isolates, and resolves network faults. The SMT Management Protocols
consist of six basic frame types; NIF, SIF, DCF, RDF, PMF, and SRF. Table A-1
describes each SMT frame type.
Table A-1. SMT Frame Types
Frame Type
Description
Neighbor Information Frames (NIF)
NIFs transmit once every 2-30 seconds so every MAC
can determine its upstream neighbor’s MAC address.
NIFs are also used to detect duplicate MAC addresses
on the ring.
Status Information Frames (SIF)
SIFs request and receive basic status information for
an FDDI station.
Echo Frames (ECF)
ECFs are used for SMT to SMT loopback (echo) testing
between FDDI stations. This determines that a
station’s Port, MAC, and SMT are operational.
Request Denied Frames (RDF)
RDFs notify SMT of illegal or inappropriate request
frame formats and protocols.
Parameter Management Frames
(PMF)
PMFs provide remote access to stations using the
get/set capabilities of the SMT MIB.
Status Reporting Frames (SRF)
SRFs use the Status Report Protocol to notify FDDI
managers of events and conditions that occur on an
FDDI station.
Connection Management
Connection Management is responsible for the insertion, removal, and connection
of the Port Physical Layer (PHY) entities to the MAC Layer entities. Connection
Management has three sub-entities that perform the tasks outlined below.
Entity Coordination Management (ECM)
ECM performs the following functions:
•
Controls the insertion and de-insertion of a station onto the FDDI ring.
•
Controls the optical bypass switch.
•
Responsible for all FDDI self diagnostics.
Each FDDI station has only one instance of ECM.
A-4
STATION MANAGEMENT (SMT)
Physical Connection Management (PCM)
PCM performs the following functions:
•
Controls the physical connection (link) of the station onto the FDDI ring.
•
Runs a line-state communications protocol between its PHY and the PHY at
the other end of the link. The line-state communications protocol tests the
integrity of the link, and checks for valid FDDI topology connections before it
allows the link to become active.
•
Checks for excessive bit errors, or line-state signals from the connected PHY
indicating errors once the link is active.
Each PHY has only one instance of PCM.
Connection Control Element (CCE)
CCE performs the following functions:
•
Controls the physical connections within a PHY.
•
Controls whether the PHY’s receive and transmit ports connect to the Primary,
Secondary, or Local ring.
Each PHY has only one instance of CCE.
Ring Management (RMT)
Ring Management controls the low level MAC Layer functions and detects MAC
Layer faults. It also performs the following functions:
•
Initiates the Target Token Rotation Time (TTRT) bidding process.
•
Watches for duplicate MAC addresses on a ring.
•
Reports missed token errors.
•
Controls where the token path is placed relative to the MAC.
A-5
ANSI STANDARDS FOR FDDI
MEDIA ACCESS CONTROL (MAC)
The FDDI Media Access Control (MAC) specifies the lower sublayer functions of
the Data Link Layer of the OSI Model. The MAC performs the following
functions:
•
Controls access to the medium (single mode fiber, multimode fiber, shielded
twisted pair, unshielded twisted pair).
•
Addresses frames.
•
Specifies token and MAC frame formats.
•
Generates MAC frames.
The MAC entity is the lower sublayer of the Data Link Layer. The upper sublayer,
Logical Link Control (LLC) serves as an interface between the OSI model and the
FDDI network. The MAC element, under control of Station Management,
performs many of the tasks associated with frame preparation and media access:
ring scheduling, initialization, and beaconing. Other tasks for the MAC entity
include assembling data frames, maintaining medium addressing, and generating
and checking data check bytes.
The MAC generates two basic message formats, tokens and frames.
PHYSICAL LAYER PROTOCOL (PHY)
The FDDI Physical Layer Protocol (PHY) specifies the upper sublayer functions of
the Data Link Layer of the OSI Model. The PHY performs the following functions:
•
Converts symbols from the MAC (encoded NRZ code bits) to the PMD
(decoded NRZI code bits).
•
Encodes data from the MAC/decodes data using a 4-bit/5-bit Encoding
Scheme.
•
Establishes clock requirements.
The following sections describe PHY functions.
A-6
PHYSICAL LAYER PROTOCOL (PHY)
4-Bit/5-Bit Encoding/Decoding Scheme
The PHY receives data frames from the MAC as a series of 4-bit symbols and
encodes each 4-bit MAC symbol as a 5-bit symbol for transmission. The 5-bit
symbols are encoded so that each symbol has at least two bit transitions to assure
bit-cell synchronization at the remote receiver. Decoding reverses this process for
the received frames. This process is referred to as 4B/5B or NRZI (Non-Return to
Zero Invert on Ones) encoding/decoding.
Other functions of the PHY include generation of a 125 Mhz transmit clock,
synchronization of the receive clock with an upstream transmitter, encoding and
decoding for media control symbols, and in some applications, buffering for the
incoming bit stream.
Clock Synchronization
The receive clock is used to recover the timing information from the incoming
serial bit stream. It is locked in frequency and phase to the local fixed frequency
oscillator.
The frequency difference between the incoming bit frequency and the outgoing
bit frequency is at the most, equal to 0.01% of the nominal frequency. The
incoming frequency can be either slower or faster than the outgoing frequency,
resulting in either an excess or deficiency of bits unless some compensation is
included.
Elasticity Buffer
An elasticity buffer is used to compensate for the difference in the frequencies. To
allow for bits that are to be dropped when the outgoing frequency is less than the
incoming frequency, the MAC entity, which originates a frame, inserts at least
sixteen IDLE symbols before each frame to be transmitted (commonly known as
PREAMBLE). The operation of the Elasticity Buffer in subsequent repeating
stations may change the length of the IDLE pattern.
An elasticity buffer is similar in function to a FIFO memory (First In First Out),
which is filled halfway before bits are removed. The input clock to the elasticity
buffer is the clock recovered from the incoming data stream. The output clock to
the elasticity buffer is the local fixed frequency oscillator for this particular
station. The minimum required elasticity is ± 4.5code bits.
The required elasticity is calculated as follows:
1.) Maximum Frame size is 9000 symbols
2.) 9000 Symbols = 4500 Code Bits
3.) Difference between Transmit and receive frequencies = 0.01%
4.) 0.01% of 45000 is 4.5 bits
5.) ± 4.5 bits is a total of 9 bits
A-7
ANSI STANDARDS FOR FDDI
PHYSICAL MEDIUM DEPENDANT (PMD)
The FDDI Physical Medium Dependant (PMD) specifies the lower sublayer
functions of the Physical Layer of the OSI model. The PMD establishes the
physical interface to the FDDI ring and converts optical energy symbols into
electrical symbols, as well as electrical energy symbols into optical energy
symbols. The PMD performs the following functions:
•
Controls optical transmit/receive levels.
•
Controls optical jitter.
•
Controls acceptable Bit Error Rates (BER).
•
Determines the fiber optic cable type.
•
Determines the connector types and pinouts.
The original PMD was designed for use with multimode fiber optic cable, but
additional PMDs are now being considered by the ANSI X3T9.5 subcommittee.
These additional PMDs include Single Mode Fiber (SMF-PMD), Low Cost Fiber
(LCF-PMD), and Twisted Pair (TP-PMD).
A-8
Appendix B
FDDI SPECIFICATIONS
This appendix outlines FDDI specifications and design considerations.
Table B-1. General Rules and Specifications
Max. Number of Connections
1000 (500 sations)
Stations are connected in series on an
optical fiber ring. Since fiber optics is
a point to point media, no taps are
allowed between stations.
Data Rate
100 Megabits per second
Max. Total Ring Length
100km (or 200 km in wrap state)
Drive Length (Max. Distance
between Stations)
-Multimode Fiber: 2 km (1.2 Miles)
-Single Mode Fiber: 60 Km (36 Miles)
-Category 5 Shielded Twisted Pair
cable: 100 Meters (328 Feet)
-Category 5 Unshielded Twisted Pair
cable: 100 Meters (328 Feet)
Transmission Media:
Fiber Optics
Multimode Fiber (MMF-PMD) as
defined by ANSI X3.166-1990.
Single Mode Fiber (SMF-PMD) as
defined by ANSI X3.184-1993.
Proposed Twisted Pair
Unshielded Twisted Pair (UTP)
Shielded Twisted Pair (STP)
Link Budget
≤ 11dB
B-1
FDDI SPECIFICATIONS
FDDI DESIGN CONSIDERATIONS
The main variables that are of interest to the FDDI network designer are:
•
ring length
•
drive distance (distance between nodes)
•
maximum number of stations on the ring
The following sections outline basic FDDI design considerations as well as critical
specifications.
Ring Length
The maximum FDDI Ring Length is 100 km. Although ANSI standard X3T9.5
does not specify ring length, it defines design parameters that are based on a total
fiber path length of 200 km.
To translate fiber path to ring length, consider that there are two counter-rotating
rings in an FDDI network. Under normal conditions (no failed segment), the ring
length is the same as the fiber path length, but if a wrap occurs, the length of the
fiber path length could nearly double. So a safe formula to establish the maximum
ring length is to divide the fiber length by two. This yields a maximum ring
length 100 km (one-half of the 200 km fiber path length). When designing a
network, add the lengths of the fiber optic cables in the network to determine the
total ring length. This includes main ring cables and branch cables that reach from
concentrators to Single Attachment Stations.
Drive Distance
Drive distance is the limit of reliable signal propagation around the ring. It is the
greatest distance that a signal can travel on the ring and still be reliably received.
For FDDI networks using fiber as defined by the FDDI standard, the maximum
drive distance is 2 km. To the network designer, this means that the maximum
cable length between any two network nodes must not exceed the 2 km drive
distance limit. In some applications, existing 50/125 µm or 100/140 µm fiber can
be used over shorter distances, but when used, the cable must conform to the
FDDI standard for bandwidth and attenuation to remain compliant with the
FDDI standard.
B-2
FDDI DESIGN CONSIDERATIONS
Attenuation
Attenuation is the level of optical power loss measured in decibels (dB). The
maximum attenuation (attenuation budget) between any two active connections
to the ring, as defined by the FDDI standard, is 11 dB. The budget includes the
attenuation of the cabling, splices, connections, and optical bypass switches.
For example, the attenuation of the typical multimode fiber optic cable used in
FDDI networks is 2.5 dB/1km or 5 dB for the 2 km maximum node separation.
The attenuation of the typical optical bypass switch is 2.5 dB. With an 11 dB
budget to work with, and 5 dB expended on the cable, the maximum number of
bypass switches is two.
Number of Stations
The maximum number of devices in a single FDDI ring is 500. This limit is
determined by the propagation delay from 1000 physical connections. With the
exception of optical bypass switches, all FDDI devices are counted as two
connections against the 1000 physical connection budget. It is easy to see how
connections are counted when only dual attached stations are used (1000 divided
by 2 connections for each Dual Attachment Stations = 500 nodes), but to
understand how connections are counted for other device types, refer to
Figure 4-5. A Dual Attachment Concentrator without attached devices is counted
as two connections (main ring connections), the same as a Dual Attachment
Station. As each Single Attachment Station or Single Attachment Concentrator is
attached to the Dual Attachment Concentrator, two connections must be counted
against the budget, one for the concentrator port and one for the attached device.
This same logic applies to counting connections for a Single Attachment
Concentrator. The multiple ports of the concentrator are not counted until a
device is attached.
B-3
FDDI SPECIFICATIONS
DAS
2
16 PHYSICAL CONNECTIONS
DAC
1
+1
2
2
SAC
2
2
SAS
2
SAS
2
SAS
2
SAS
2
SAS
Figure B-1. Physical Device Connections
B-4